We developed an assay that subdivides cell adhesion into at least two events, an initial binding event we refer to as 'affinity' and a secondary step or steps that stabilize the interaction. The initial binding step has some specificity properties and can help explain a number of phenomena that have been described but little understood in cell aggregation studies. The assay has been modified to screen monoclonal antibodies directly for inhibition of the initial affinity step between cells. A large number of retinal cell surface antibodies have been screened with the assay. Nine monoclonals identify eight different molecules that inhibit cell recognition. Two of the monoclonals identify nCAM, providing an independent confirmation that this retina cell surface molecule is involved in cell interactions. This proposal includes experiments to further characterize the antigens identified, and to begin to understand how multiple components cooperate in the initial steps of cell recognition during development. Three parameters will be focal points for these studies. First, with multiple antigens contributing to the same interaction, it becomes difficult to detect 'minor' components. A model is developed for studying such multi-component systems and the predictions of such a model will be tested. Second, specificity in the retina will be studied using a dorsal-ventral gradient that can be observed with great repeatibility using the cell binding assay. Experiments that affect the D-V specificity ratio will attempt to establish its molecular basis and then its function. Third, the retina is a heterogeneous tissue made of many cell types. We will take advantage of cell sorting, and other protocols to isolate individual cell populations to ask how they interact with other retina cells. Finally, we will pursue approaches that allow us to examine the surface of the optic tectum, the site where the dorsal-ventral gradient or retinotopic specificity is manifest. We will develop methods for looking only at the surface since evidence suggests that other cells of the tectum do not share the properties of the gradient that is present there.